Rjeas Research Journal in Engineering and Applied Sciences 1(5) 338-341 Rjeas

© Emerging Academy Resources (2012) (ISSN: 2276-8467)

www.emergingresource.org

A SMOKER EQUATION BASED ALGORITHM FOR BINARY DISTILLATION

DESIGN CALCULATIONS

Osoka, E.C. and Kamalu, C.I.O.

Department of Chemical Engineering
Federal University of Technology
P.M.B. 1526, Owerri.
Corresponding Author: Osoka, E.C
___________________________________________________________________________
ABSTRACT
An algorithm- based on smoker’s analytical equation- suitable for the development of computer programs, for
the determination of some design parameters of a binary distillation column was presented. A sample problem
was modified, with respect to constant relative volatility, and solved using the algorithm. We compared the
solution obtained using the algorithm to that using McCabe-Thiele method. The deviation of 4.25% observed
can be explained based on the assumption of constant relative volatility. This deviation is not only acceptable in
comparison to a deviation of about 23.25% that could exist between solutions obtained using McMabe-Thiele
and Fenske’s method at total reflux, but also safe since it does not reduce but adds to the number of theoretical
plates, thus ensuring that operational targets of purity in the product are met. Apart from suitability under
normal conditions, the algorithm was designed for use when concentration of more volatile component of either
product is very low or relative volatility is low, conditions under which the McCabe-Thiele method is
impracticable.
©Emerging Academy Resources

KEYWORDS: Binary Distillation, Mccabe-Thiele Method, Smoker’s Equation, CAD, Relative Volatility,
Lewis-Sorrel Method.
________________________________________________________________________________________
INTRODUCTION
Distillation is agreeably the most widely applied
method of separation because it has the advantage of
handling a wide range of throughput, feed
composition and also the ability to achieve high
product purity. Binary distillation is a unit operation stages for a distillation column operating at a nearly
for separating two components in a mixture between
a vapour and liquid phase based on the difference in
the volatilities of the components (Sinnot, 2005;
Richardson et. al., 2003). In calculating the design
parameters for binary distillation, two main methods
are applied. These are Lewis-Sorrel method and
McCabe-Thiele method. The Lewis-Sorrel method
can also be applied to multi-component distillation to
determine the number of plates and is the basis of
modern computerized methods for design of
distillation columns (Treybal, 1982; Perry and Green,
1997; Luyben, 2008).
Techniques like McCabe-Thiele method become
impracticable when concentration of more volatile
component of either product is very low or the
volatility is low.
Alleva (1962) offered a solution to the first problem
by using a log-log plot, while Robinson and Gilliland
(1950) presented two equations for obtaining the
number of plates for the rectifying and stripping
sections.
Smoker (1938) presented an analytical equation that
can be used to determine the number of equilibrium
constant relative volatility. Though this method is
suitable for any problem in which the relative
volatility for the rectifying and stripping sections may
be taken as being relatively constant, it is particularly
useful when the relative volatility is low or
concentration of more volatile component in either
product is very low, thus making it robust.
This study presents a technique that takes care of
situations when McCabe-Thiele diagram is
impracticable, like the case of separation of close
boiling isomers, where the relative volatility is close
to unity.
Modern engineering practice is becoming largely
dependent on computer and information technology
due to the phenomenal growth in computer hardware.
Hence the design, maintenance and operation of
plants are performed using computer-aided design
(CAD) (Teicholz, 1985; Cogner, 1994; Chandrapatla
and Berry, 1996).

338
 Research Journal in Engineering and Applied Sciences (ISSN: 2276-8467) 1(5):338-341
A Smoker Equation Based Algorithm for Binary Distillation Design Calculations

There is therefore an ever-increasing need to modify If the feed stream is not at its boiling point, xf - for
CAD techniques using methods found to be the stripping section - is replaced by the value of x at
computationally efficient and robust the intersection of the operating lines;
* b  ( x f /(q  1))
Derivation of Equations xf  (16)
A straight operating line can be represented by the (q /(q  1))  S
equation;
y  Sx  b (1) Algorithm for Binary Distillation Design
In terms of relative volatility, the equilibrium values Calculations
of y are given by; Input: α, R, xd, xf, xb, q
Rectifying Section Calculations
x
y (2) E10  S
1  (  1) x
Eliminating y from equations (1) and (2) gives a E11  b
quadratic equation in x of the form; E4  k , subject to 0 < k < 1
S (  1) x 2  ( S  (  1)   ) x  b  0 (3) E8  x0
*

For any particular distillation problem, the above *

equation will have only one real root, k, between 0 E9  xn
and 1, such that;
E7  c
S (  1)k 2  (S  (  1)   )k  b  0 (4)
Where k is the value of the x co-ordinate at the point
E6  
where the extended operating lines intersect with the E5  N R
vapour-liquid equilibrium curve. Stripping Section Calculations
Smoker shows that the number of stages required is If feed is not at boiling point
given by the equation;
E16  new x f
 x * (1   xn* ) 
Ni  log o* * 
 / log ( / S c 2 ) (5) Else; retain x f
 xn (1   xn ) 
E14  S
Where,
S c (  1) E15  b
 (6)
  Sc
2 E4  k , subject to 0 < k < 1
*
Ni is the number of stages required to effect the E12  x0
separation represented by the concentration change *
from x0* to xn*. E13  xn
Where x0*= (x - k) and x0* > xn*. E7  c
c  1  (  1)k (7)
S = slope of the operating line E6  
α = relative volatility, assumed constant over x0* to E5  N S
xn*.
Compute N = NR + NS
For a column with a single feed and no side streams,
rectifying section equations will be given as;
A Sample Calculation
*
x0  xd  k (8) A continuous rectifying column treats a mixture
* consisting of 40% of benzene by weight and 60% of
xn  x f  k (9) toluene at the rate of 4kg/s, and separates it into a
S  R /( R  1) (10) product containing 97% of benzene and a liquid
containing 98% toluene. The feed is at its boiling
b  xd /( R  1) (11) point.
Stripping section equations will be given as;
* If a reflux ratio of 3.5 to 1 is employed, how many
x0  x f  k (12)
plates are required in the column (Richardson et. al.,
*
xn  xb  k (13) 2003).

R x f  xd  ( R  1)k  xb Solution
S (14) We modify the problem by using a relative volatility
( R  1)( x f  xb ) of 2.46 instead of the equilibrium data (Richardson
( x f  xb ) xb et. al., 2003). This value of relative volatility closely
b (15) matches with the equilibrium data.
( R  1)( x f  xb )

339
 Research Journal in Engineering and Applied Sciences (ISSN: 2276-8467) 1(5):338-341
A Smoker Equation Based Algorithm for Binary Distillation Design Calculations

Input: α = 2.4600, R = 3.5000, xd = 0.9740, xf = The Fenske’s equation is another relation that
0.4400, xb = 0.0240. assumes a constant relative volatility. Backhurst and
Harker (1999) compared calculation of number of
The input data were obtained by converting the plates at total reflux, using both McCabe-Thiele and
values given as percentage by weight to their molar Fenske’s equation, and concluded that a deviation
equivalents (Backhurst and Harker, 1999). (reduction in actual number of plates required) of
about 23.25% was justifiable based on the
Output 1: S = 0.7778 assumption of constant relative volatility.
b = 0.2164
k = 0.1877 CONCLUSION
x0* = 0.7863 In comparison to the deviation between Fenske’s
xn* = 0.2523 equation and McCabe-Thiele method, the solution
c = 1.2740 obtained using the algorithm based on Smoker’s
β = 1.2081 analytical equation is acceptable. It can also be
NR = 5.6512 observed that the deviation adds to the number of
Output 2: S = 1.2853 theoretical plates obtained using McCabe-Thiele
b = -6.8462 x 10-3 method, thus ensuring the achievement of the
k = 0.6370 required product purity, unlike a deviation that
x0* = -0.1970 reduces the actual number of theoretical plates, as
xn* = -0.6130 was the case of Fenske’s equation. Since the
c = 1.9300 algorithm computes the number of plates in the
β = -1.5560 rectifying and stripping sections of the column
NR = 5.7736 separately, it also gives an indication of the position
N = 11.4248 of the feed plate.

DISCUSSION The algorithm is therefore a robust tool for binary

Backhurst and Harker (1999) solved the same
problem presented in this work using the McCabe-
Thiele method and obtained the solution as ten (10)
theoretical plates. In consideration that the reboiler
and condenser account for one theoretical plate, the
number of theoretical plates based on our algorithm
becomes 10.4248 theoretical plates. The observed
deviation (a deviation of 4.25%) can be explained
based on the assumption of constant relative volatility
and its use in place of the actual equilibrium data.
distillation column design calculations, both under
normal conditions and those of low volatility, or
when concentration of more volatile component in
either product is very low.
However, this technique cannot be applied for cases
where the relative volatility varies widely and cannot
be adequately represented with an average value.

NOMENCLATURE
Symbol Definition Units
Ei Equation (i) Nil
x Molar fraction of more volatile component Dimensionless
xf Molar fraction of more volatile component in feed Dimensionless
xd Molar fraction of more volatile component in distillate Dimensionless
xb Molar fraction of more volatile component in bottoms Dimensionless
α Relative volatility Dimensionless
R Reflux ratio Dimensionless
NR Number of plates in the rectifying section Dimensionless
NS Number of plates in the rectifying section Dimensionless
N Number of plates in the rectifying section Dimensionless
q Heat required to vapourize one mole of a component J/mol

REFERENCES Chandrapatla, T. R., and K. J. Berry (1996)

Alleva, R.Q. (1962) “Improving McCabe-Thiele “Advances in Engineering Software: including
diagram”, Chem. Engr., Vol. 64 (August 6th), New computing systems in Engineering”, Elsevier Science
York, pp. 111. Limited, USA, Vol. 7, no. 2, pp. 191-199.

Backhurst, J. R., and J. R. Harker (1999) “Coulson Cogners, S. A. (1994) “The New Software
and Richardson’s Chemical Engineering”, Vol. 5, 2nd Engineering”, Wadsworth publishing company,
ed., Butterworth-Heinemann, Oxford, London, pp. Belmont, California, pp. 501-539
109-110,137.

340
 Research Journal in Engineering and Applied Sciences (ISSN: 2276-8467) 1(5):338-341
A Smoker Equation Based Algorithm for Binary Distillation Design Calculations

Luyben, W. L. (2008) “Process Modelling,

Simulation and Control for Chemical Engineers”, 2nd
ed., McGraw-Hill, Singapore, pp. 8-9, 69-75.

Perry, R. H., and D. W. Green (2008) “Perry’s

Chemical Engineering Handbook”, 8th ed., McGraw-
Hill, New York, pp. 13-22.

Richarson, J. F., Harker, J. H., and J. R. Backhurst

(2003) “Coulson and Richardson’s Chemical
Engineering”, Vol. 2, 5th ed., Butterworth-
Heinemann, Oxford, London, pp. 542-
560,1164,1165.

Robinson, C. S., and E. R. Gilliland (1950)

“Elements of Fractional Distillation”, McGraw-Hill
Company, London.

Sinnot, R. K. (2005) “Coulson and Richardson’s

Chemical Engineering”, Vol. 6, 2nd ed., Butterworth-
Heinemann, Oxford, London, pp. 456-462.

Smoker, E. H. (1938) “Analytical Determination of

Plates in a Fractionating Column”, Trans. Am. Inst.
Chem. Engr., Vol. 34, pp. 165.

Teicholz, E. E. (1985) “CAD/CAM Handbook”,

McGraw-Hill Inc., USA, pp. 210-213.

Treybal, R. E. (1982) “Mass Transfer Operations”, 3rd

ed., McGraw-Hill Inc., USA, pp. 423

341